Expert Experience Research Articles

Differences in Perspectives Between Experts and Residents on Living Heritage: A Study of Traditional Chinese Villages in the Luzhong Region

The differences in the heritage perspectives of key stakeholders form an indispensable basis for formulating appropriate conservation strategies for living heritage. However, in existing practices, the differences in heritage perspectives between experts and local residents, who are both key stakeholders, often arise from expert experiences rather than being accurately measured. This study regarded traditional Chinese villages in the Luzhong region as a living heritage case for investigation and quantified such differences. Initially, this study provided a comprehensive description of living heritage using four dimensions: heritage value, heritage attributes, heritage composition, and heritage characteristics. Subsequently, a questionnaire tailored to traditional Chinese villages was developed, and a survey was conducted in the sample villages, resulting in 394 valid responses. According to the Pearson chi-square tests, the findings suggest that expert and local resident respondents showed slight disagreement regarding heritage attributes (x2 = 5.619); however, they presented varying degrees of differences concerning heritage value (x2 = 36.066) and heritage composition (x2 = 36.525), accompanied by distinct preferences. Furthermore, significant discrepancies were evident regarding heritage characteristics. Regarding the four aspects of heritage characteristics, there was a slight difference in the symbiosis of heritage elements (x2 = 3.877) but significant differences in the local resident as a heritage element (x2 = 36.525); there was a minor difference in vitality and continuity (x2 = 3.709) but a rare contradiction for integrity (x2 = 47.649). This study can furnish case data support for surmounting the drawbacks of blindly relying on expert experiences, particularly by integrating local perspectives to safeguard living heritage globally.

Development of Preliminary Criteria of Macrophage Activation Syndrome in Multisystem Inflammatory Syndrome Associated with COVID-19 in Children

Background: Macrophage activation syndrome (MAS) can be regarded as a key factor determining the severity of multisystem inflammatory syndrome associated with COVID-19 in children (MIS-C), and often requires treatment in the intensive care unit (ICU) to avoid life-threatening complications. No reputable specific criteria for the diagnosis of MAS in MIS-C patients have yet been identified, and criteria currently used for the diagnosis of hemophagocytic syndromes, such as HLH-2004, MAS-2005, and MAS-2016, are not sufficient for MAS in MIS-C. Our goal in this study was to work out the criteria for the early diagnosis of MAS in MIS-C. Methods: One hundred and sixty-six (166) patients with MIS-C were assessed retrospectively. The two most experienced experts independently identified patients with MAS. The patients were divided into three cohorts: MAS (n = 19), without MAS (n = 78), and probable MAS (n = 67). The latter included patients diagnosed with MAS by only one expert, and it was excluded from the analysis. Results: The age of patients with MAS was much higher, and they more frequently had edematous syndrome, hypotension and/or shock, splenomegaly, and CNS involvement. In their blood tests, thrombocytopenia, hypoalbuminemia, and hypertriglyceridemia occurred more often. The level of biomarkers of inflammation, such as ferritin, CRP, troponin, AST, and ALT, was also higher in this group. Increased fibrinogen and D-dimer were also found, demonstrating hypercoagulation in the MAS-MIS-C group. We chose 21 continuous and categorical variables with statistical significance, out of which 2—ferritin > 469 μg/L or platelets < 114 × 109/L—allowed us to discriminate MAS patients. Conclusions: Ferritin > 469 μg/L or platelets < 114 × 109/L can be regarded as key signs to differentiate MAS in MIS-C patients with a sensitivity of 100% and specificity of 94.9%, and they can be used along with other diagnostic methods.

Temperature Control Strategy for Hydrogen Fuel Cell Based on IPSO-Fuzzy-PID

Hydrogen fuel cell water-thermal management systems suffer from slow response time, system vibration, and large temperature fluctuations of load current changes. In this paper, Logistic chaotic mapping, adaptively adjusted inertia weight and asymmetric learning factors are integrated to enhance the particle swarm optimization (PSO) algorithm and combine it with fuzzy control to propose an innovative improved particle swarm optimization-Fuzzy control strategy. The use of chaotic mapping to initialize the particle population effectively enhances the variety within the population, which subsequently improves the ability to search globally and prevents the algorithm from converging to a local optimum solution prematurely; by improving the parameters of learning coefficients and inertia weight, the global and local search abilities are balanced at different stages of the algorithm, so as to strengthen the algorithm’s convergence certainty while reducing the dependency on expert experience in fuzzy control. In this article, a fuel cell experimental platform is constructed to confirm the validity and efficiency of the recommended strategy, and the analysis reveals that the improved particle swarm optimization (IPSO) algorithm demonstrates better convergence performance than the standard PSO algorithm. The IPSO-Fuzzy-PID management approach is capable of providing a swift response and significantly diminishing the overshoot in the system’s performance, to maintain the system’s safe and stable execution.

Decision Making in the Searching Weakness Process of Server Side Web Application.

The vulnerability search process consists of three stages. At the first stage, the weakness of software and firmware components is identified and their suitability for exploitation. The second stage is devoted to choosing an existing or developing a new exploit for the identified weakness. The third stage configures the usability and reliability of the exploit for the identified vulnerability. The vulnerability of the system of the cyber protection object to negative technical impact is always based on the exploitation of its defects. Weakness identification with applicability assessment is a multi-stage process of choosing solutions under conditions of uncertainty. The efficiency of the weakness search process can be improved by using decision making theory. The idea is to reduce uncertainty by using an information model to search for weakness and to make it possible to process it quantitative analysis. For quantitative analysis, classical criteria for choosing solutions under uncertainty conditions were used. The information model is a hierarchy, has four levels and includes four types of elements. First level consists a techniques of Initial Access. First level consists a techniques of Initial Access, second – of category wekness Web Application. The third and fourth levels of assessment, respectively, of weaknesses and vulnerabilities. The expediency, rationality and reasonableness of solutions for finding defects is based on the application of proven sources of world-class expert experience. Thus, Adversarial Tactics, Techniques & Common Knowledge (ATT&CK) defines Server-side Web Application class systems as an attractive attack target for hackers with the highest average number of potential defects in at least three components: Web Server, Web Application Server, DBMS Server. To find out the distribution of Server-side Web Application defects by characteristic classes, it is necessary to use the Open Worldwide Application Security Project (OWASP), for a detailed study of them - Common Weakness Enumeration (CWE). The application of the National Vulnerability Database (NVD) and Common Vulnerabilities and Exposures (CVE) complements the evaluation of the found defect.

Addressing the necessity of cybersecurity literacy: the case of ETTCS CyberTeach project

Purpose The purpose of this study is to present actions and initiatives, developed within the scope of the Empowering Teachers to Trigger Cybersecurity at Schools (ETTCS) CyberTeach Erasmus Project, to enhance cybersecurity literacy through innovative digital content and contemporary Learning Management System (LMS) platform. Design/methodology/approach The concept of cybersecurity literacy has become increasingly crucial in recent years, as the digitization of most human activities is being completed within the framework of the 4th Industrial Revolution. Almost all devices, vehicles and services in the near future will be interconnected to the internet and operate on advanced computing platforms. The benefits of these technological advancements are evident, as are the potential risks. To this end, organized cybercriminals, black hat hackers and state-sponsored actors may attempt, through various cyberattacks, to steal personal data, cause road accidents in connected autonomous vehicles and, in general, disrupt critical infrastructures. Cybersecurity is a growing concern when it comes to digitalization and cloudification. This way, digital assets must be conveniently protected to avoid any concern about their confidentiality, integrity and authentication. Therefore, the ability of every citizen to use the internet and smart devices wisely and securely is one of the most fundamental skills they should possess. In this work, the authors present actions and initiatives, developed within the scope of the ETTCS CyberTeach Erasmus Project, to enhance cybersecurity literacy through innovative digital content and contemporary LMS platform. A new approach to teach cybersecurity, based on innovative teaching methods, is presented to prepare future citizens and their teachers to keep up with cybersecurity issues in an efficient manner. To this end, the authors propose ways to reach cybersecurity literacy, giving use case examples and proposing the necessary digital skills. Findings A new approach to teach cybersecurity, based on innovative teaching methods, is presented to prepare future citizens and their teachers to keep up with cybersecurity issues in an efficient manner. To this end, the authors propose ways to reach cybersecurity literacy, giving use case examples and proposing the necessary digital skills. Originality/value This work demonstrates a new methodology to infuse cybersecurity awareness into teachers so that they can train and prepare their students accordingly. The authors identify the teacher as the critical link between a young generation of digital natives, who consider the internet as a utility, and the most experienced information technology security experts striving to enforce good cybersecurity practices among users.

Editor's view: Is the brain's perception of ideas an underappreciated human sense?

Humans have developed sensory organs for their 'major' senses and specific neuronal receptors for a number of additional senses. Although information from sensors is integrated and processed in the brain, the brain itself has not been proposed as a sensory organ associated with a particular sense - at least not in Western culture or scientific literature. Perception of ideas has many elements of a separate sense, with the brain being a primary sensory organ. I support this notion based on 17 years of experience in the application of the Child Health and Nutrition Research Initiative (CHNRI) method for setting research priorities, which resulted in more than 150 publications involving thousands of experts who prioritised more than 20 000 research ideas. This body of work allows for the generalisation of the following key concepts: defining an idea in the context of the brain's sensory perception; defining the point of origin, sensory qualities, and common types of ideas; considering responses of the body to the brain's sensory exposure to ideas; and defining the key criteria that the brain uses to discriminate between and prioritise ideas. The human brain is continuously being exposed to ideas, which can be self-generated, triggered by information from other sensors, or introduced from the external world. From the brain's sensory perspective, ideas are competing possibilities of purposeful activities that, if followed, would be expected to result in an alternative version of the future. Pursuing ideas tends to drive most of human activity. It requires prioritising between short-, mid-, and long-term investment of energy and time. The brain's sensory role is to continuously assess many competing ideas and prioritise between them based on motivational/emotional ('attractiveness'), operational/rational ('feasibility') and outcome-related perspective ('impact'). Exposure to new ideas may instigate physiological and psychological responses, ranging from enthusiasm and excitement to feeling a threat or fear. Ideas can be used to mobilise large groups of people whose brains respond with excessive enthusiasm that can sometimes be fanatical. The judgment on the attractiveness, feasibility, and potential impact of ideas is affected by education, experience, and cognitive abilities. This sense may be sharpened through an increased level of expert knowledge and experience. Misinformation and disinformation are dangerous because they affect the brain's perception of ideas. Impairment of this inherent sense may perhaps contribute to some mental health issues. The proposed view of the brain as the sensor of ideas could lead to many qualitative or quantitative experiments to further explore the properties of this sense in both individuals and populations, establishing 'the science of ideas'.

Data imbalanced fault diagnosis of gearbox transmission system under various speeds based on dynamic dual-scale normalized fusion network

Abstract Gearboxes play a pivotal role in industrial production, and their reliability and safety are essential for production safety and efficiency. However, gearboxes frequently encounter challenges such as variable rotational speeds and unknown operating conditions. Unfortunately, most existing traditional fault diagnosis methods face the following issues: (1) They heavily rely on expert experience and pre-existing knowledge bases, making them unable to tackle fault diagnosis in unknown working conditions. (2) While addressing various speed issues, they seldom consider the problem of data imbalance in real-world industrial environments. (3) Many transfer learning methods primarily focus on global distribution alignment and knowledge transfer between source and target domains, neglecting the importance of fine-grained distribution alignment between subdomains. To address these issues, a dynamic dual-scale normalization fusion network (DSNFNet) is proposed for cross-domain fault diagnosis under variable speed and data imbalance. Firstly, the two parallel graph convolution frameworks constructed are used to extract multi-scale fault features. Subsequently, a dual-scale normalization fusion module is adopted to integrate the global and local fault feature information within the subdomains of both the source and target domains, thereby aligning their fine-grained distributions to obtain domain-invariant features. Finally, a dynamic soft threshold feedback strategy is introduced, which assigns pseudo labels to the target domain data, enabling the model to give equal attention to each class of data samples, even under data imbalance conditions, thereby improving the recognition accuracy of minority fault classes. Validating the proposed method on two real cases, our method achieved the highest accuracy compared to other advanced approaches.

Endocrine management of transgender and gender diverse adolescents: expert opinion of the ESPE Working Group on Gender Incongruence and the Endo-ERN main thematic group on Sexual Development and Maturation.

Increasing numbers of transgender and gender diverse adolescents seek endocrine treatment to align their body to their gender identity. In this relatively young field of medicine, there is a limited body of evidence, and the available evidence generally is of low quality. However, in the absence of randomised trials, much can be learned from prospective observational studies. While evidence on somatic and psychosocial outcomes has increased in recent years, areas with little or no research rely on expert opinion. In order to reduce practice variation in paediatric endocrine care for transgender adolescents, we describe endocrine care for transgender and gender diverse adolescents in Europe in this statement, based on the available evidence as well as the experience of experts from the ESPE Working Group on Gender Incongruence and from the European Reference Network Endo-ERN, main thematic group 'Sexual Development and Maturation'. This document aims to provide practical information and tools for clinical care, with sections on the multidisciplinary team, counseling, fertility, hormone suppression, gender affirming hormone treatment, support for non-binary adolescents, long-term follow up and transition of care, altered treatment wishes, lifestyle and sexual health, and in addition includes a paragraph on ethical considerations regarding this care.

Unsupervised machine learning in financial anomaly detection: clustering algorithms vs. dedicated methods

The article presents the application of selected clustering algorithms for detecting anomalies in financial data compared to several dedicated algorithms for this problem. To apply clustering algorithms for anomaly detection, the Determine Abnormal Clusters Algorithm (DACA) was developed and implemented. This parameterized script (DACA) allows clusters containing anomalies to be automatically detected on the basis of defined distance measures. This kind of operation allows clustering algorithms to be quickly and efficiently adapted to anomaly detection. The prepared test environment has allowed for the comparison of selected clustering algorithms. K-Means, Hierarchical Cluster Analysis, K-Medoids, and anomaly detection: Stochastic Outlier Selection, Isolation Forest, Elliptic Envelope. The research has been carried out on real financial data, in particular on the income declared in the asset declarations of the targeted professional group. The experience of financial experts has been used to assess anomalies. Furthermore, the results have been evaluated according to a number of popular classification and clustering measures. The highest result for the investigated financial problem was provided by the K-Medoids algorithm in combination with the DACA script. It is worthwhile to conduct future research on the introduced solutions as an ensemble method.

Integrating computing techniques in preserving Makassar’s Pakarena dance heritage

This interdisciplinary study bridges the cultural heritage of the Pakarena dance, a key element of Makassar ethnicity, with contemporary computing and informatics to examine its social implications, particularly in terms of identity formation, community transmission, and narrative preservation. Emphasizing the dance’s crucial role within the Makassar community, the research employs digital technologies for comprehensive data collection, analysis, and dissemination. Utilizing a qualitative framework supplemented by digital ethnography, the methodology includes in-depth interviews and participant observation, enriched with advanced data analysis and virtual reality (VR) presentations. This innovative approach facilitates the digital capture of the narratives and experiences of dance practitioners, cultural experts, and community members, ensuring the preservation of the dance’s cultural narrative and expression. The study reveals that the Pakarena dance is not only a bearer of the Makassar community’s history and traditions but also a platform for individual creativity and cultural identity, adapting while preserving its core in the face of societal shifts. The findings highlight the potential of computing and informatics in cultural preservation, suggesting new methods for documenting, analyzing, and promoting intangible cultural heritage. The study advocates for the use of technology to enhance and perpetuate cultural heritage, especially for younger generations, in our increasingly digital era.

Involvement of representatives of patients and relatives in guideline processes : S3 guidelines on psychosocial therapies for severe mental illnesses breaks new ground

The participation of representatives of patients and relatives in the development of guidelines is of central importance for the quality, feasibility and acceptance of guidelines. In Germany this has not been sufficiently implemented in the discipline of psychiatry, although in practice there are many examples for the benefits of the expertise of people with psychiatric experience. The article describes the development and a first process evaluation of a trialogical working group (AG Impuls) accompanying the guideline process. Possibilities for further development and implementation are discussed. Description of the working process of the Impuls working group as well as the results of the satisfaction survey after 18months and summarised content analysis. Concept: during the further development of the S3 guidelines on psychosocial therapies, the expertise of 12experts is bundled in the Impuls working group and supported by the members of the steering group. An open discussion takes place in regular digital working meetings and face to face meetings are held once ayear as part of the German Association for Psychiatry, Psychotherapy and Psychosomatics (DGPPN) congress. The results are presented to the guideline committee at the consensus conferences. Text boxes from the Impuls working group constitute an important tool. the following topics were identified in connection with the participatory work: (a)exchange at eye level, broadening perspectives and contributing impulses, (b)giving avoice to experienced experts, (c)implementation requirements and possibilities and (d)limitations of the cooperation. Various preconditions and implementation requirements for long-term successful participation in guideline development were identified. The Impuls working group can be used to specifically address insufficient participation in guideline processes. The structured approach could form ablueprint for other guideline processes.

Research on the multi-area cooperative control method for novel power systems

As new energy sources increasingly penetrate novel power systems, the coordinated control of frequency regulation units has become more challenging, leading to the degraded performance of power grid control. The automatic generation control methods based on traditional reinforcement learning overly depend on exploration, making it difficult quickly obtain optimal solutions for multi-area cooperative control and resulting in suboptimal performance. To address this issue, this paper proposes a multi-area coordinated control method using a greedy actor-critic algorithm enhanced with expert experience replay. This method decouple the role of entropy for exploration and policy collapse, utilizing a proposal policy to secure high-value actions for policy updates. Thus the exploration and exploitation are balanced and the local optima is avoided. Additionally, the expert experience replay collects expert demonstration data with high-value to assist the learning of multi-agent. Therefore, the meaningless exploration in early training is reduced and the rapid attainment of optimal solutions is facilitated. This paper validates the proposed method through simulations on the model of the improved IEEE standard two-area load frequency control and the model of the Sichuan, Chongqing, and Hubei three-area. Compared with other various reinforcement learning methods, the proposed method is demonstrated with superior control performance.

A fuzzy decision-making network model for offshore wind turbine selection based on simulated annealing algorithm

As global demand for renewable energy surges, offshore wind power has become an indispensable component in the transition towards a more sustainable energy landscape. Selecting appropriate wind turbines for offshore wind power facilities is a critical process influenced by various factors, including turbine parameters, environmental conditions, and economic indicators. Traditional decision-making methods often fall short in addressing the complexity and dynamic nature of real-world scenarios, relying heavily on expert experience. This study proposes a novel and scalable method for offshore wind turbine selection by leveraging a fuzzy decision-making network and a conditional probability table optimized through a simulated annealing algorithm. The method is structured into three key modules. It allows for learning from existing similar cases and transferring knowledge to the current selection task in the absence of expert opinion. It has proven its worth in a case study at the Rudong Wind Farm in Jiangsu Province, China. The results show that this method can account for the complexities inherent in turbine selection and serve as a more comprehensive and adaptable tool for decision-makers. This study contributes to the renewable energy sector by providing a novel, data-driven, and practical method for offshore wind turbine selection and other real-world related applications.

PERSPECTIVES OF PSYCHOLOGISTS TOWARDS THE FIELD OF ERGONOMICS: A QUALITATIVE STUDY

Employees need a healthy environment and working conditions in their workplaces. This need is a legal obligation that must be met by employers. Ergonomics is concerned with improving the compatibility between humans and their environment. Considering that humans are psychosocial beings, addressing the field of ergonomics independently of psychology can be seen as a limitation. Therefore, examining the perspectives of psychologists working in workplaces regarding the field of ergonomics is valuable. This study aims to understand the perspectives on ergonomics of psychologists from different specialties and sectors in Turkey. Fifteen experts working in various cities across Turkey participated in the research. The study is qualitative, and the perspectives of the experts were explored through semi-structured interviews. Thematic analysis was chosen as the method, and the analysis revealed three different themes regarding perspectives on the field of ergonomics: (i) psychologists' knowledge of ergonomics, (ii) psychologists' experiences with ergonomics-related work, and (iii) psychologists' perspectives on the intersection of psychology and ergonomics. The results of the study indicate that the knowledge and experiences of experts regarding ergonomics are limited. Furthermore, it was found that psychologists do not have structured written job descriptions assigned by their workplaces, leading to role conflict and role ambiguity.

Improving the Classification of Defect Levels in Main Electrical Equipment Through Advanced Prompting on a Large Language Model

The stable operation of the power system is closely related to the national economy and people’s livelihoods. Therefore, the timely detection, qualitative assessment, and handling of major equipment defects are crucial. The classification of defect levels in main electrical equipment is a fundamental task in this process, which is often manually completed, supplemented by knowledge bases or expert systems. However, this approach is time-consuming, labor-intensive, involves challenging human–machine interaction, and relies on expert experience. Conversational large language models, such as ChatGPT, ERNIE Bot, ChatGLM, have garnered widespread recognition in various domains. However, these models may have errors in the reasoning process, resulting in biased or even erroneous outputs, which is referred to as the “hallucinations”. The hallucinations’ problem of large language models poses challenges in specific fields. To mitigate the hallucinations in large language models, researchers often seek to incorporate domain-specific knowledge into these models through methods like fine-tuning or prompt learning. In order to enhance model performance while minimizing computational costs, this study adopts the prompt learning approach. Specifically, we propose a large language model prompt learning framework based on knowledge graphs, aiming to provide the large language model with reasoning support by leveraging specific information stored within the knowledge graph and receive explainable reasoning result. Experimental results demonstrate that our module achieved superior results on the power defect dataset compared to the non-prompt method.

Segmentation of infarct lesions and prognosis prediction for acute ischemic stroke using non-contrast CT scans

Background and PurposeIschemic stroke is the most common type of stroke and the second leading cause of global mortality. Prompt and accurate diagnosis is crucial for effective treatment. Non-contrast CT (NCCT) scans are commonly employed as the first-line imaging modality to identify the infarct lesion and affected brain areas, as well as to make prognostic predictions to guide the subsequent treatment planning. However, visual evaluation of infarct lesions in NCCT scans can be subjective and inconsistent due to reliance on expert experience. MethodsIn this study, we propose an automatic method using VB-Net with dual-channel inputs to segment acute infarct lesions (AIL) on NCCT scans and extract affected ASPECTS (Alberta Stroke Program Early CT Score) regions. Secondly, we establish a prediction model to distinguish reperfused patients from non-reperfused patients after treatment, based on multi-dimensional radiological features of baseline NCCT and stroke onset time. Thirdly, we create a prediction model estimating the infarct volume after a period of time, by combining NCCT infarct volume, radiological features, and surgical decision. ResultsThe median Dice coefficient of the AIL segmentation network is 0.76. Based on this, the patient triage model has an AUC of 0.837 (95 % confidence interval [CI]: 0.734–0.941), sensitivity of 0.833 (95 % CI: 0.626–0.953). The predicted follow-up infarct volume correlates strongly with the DWI ground truth, with a Pearson correlation coefficient of 0.931. ConclusionsOur proposed pipeline offers qualitative and quantitative assessment of infarct lesions based on NCCT scans, facilitating physicians in patient triage and prognosis prediction.

Development of Pupils', Students' and Cadets Value Attitude to Health-Preserving Activities by Means of Combat Horting

The co-authors of the scientific publication carried out the research and described the results by means of a survey of 9th grade athletes from 34 sports clubs in Ukraine (n=807), research students of the State Institution Luhansk Taras Shevchenko National University (n=92), and practicing students of the State Tax University (n=32), who were involved in the development of mini-projects on health promotion activities, conducting training in combat horting in natural conditions, experimental extracurricular health activities, writing qualification papers on the topic of research and participating in conferences of various levels, including international ones, monitoring activities during combat horting classes and extracurricular activities). The aim of the study is to practically test the efficiency of educating pupils', students' and cadets value attitude to health-preserving activities by means of combat horting. The basis of the study was the development of research skills through the method of analyzing student researchers; improvement of practical skills of pupils and students who participated as combat sport athletes and student practitioners, assistant coaches from 2020 to 2024. Results of the research. The co-authors of the research group developed and implemented a set of goals for educating pupils', students' and cadets to cultivate a value-based attitude to health promotion; developed questionnaires and conducted a survey among pupils and students; identified components, criteria and indicators for developing a value-based attitude of pupils', students' and cadets to health promotion; and presented practical recommendations. Moreover, the respondents acted as coaches, which promoted the development of soft skills and the ability of teamwork. Conclusions: This study contributed to the coordination of teamwork between pupils, students, teachers/coaches; aimed at improving the quality of professional and practical training of higher education students and their career development. Consequently, it is necessary to focus on the practical component of educational and research activities in higher education institutions, improvement of the educational and training process in sports clubs of Ukraine, based on the domestic and foreign experience of prominent experts.

Human-in-the-Loop Reinforcement Learning in Continuous-Action Space.

Human-in-the-loop for reinforcement learning (RL) is usually employed to overcome the challenge of sample inefficiency, in which the human expert provides advice for the agent when necessary. The current human-in-the-loop RL (HRL) results mainly focus on discrete action space. In this article, we propose a Q value-dependent policy (QDP)-based HRL (QDP-HRL) algorithm for continuous action space. Considering the cognitive costs of human monitoring, the human expert only selectively gives advice in the early stage of agent learning, where the agent implements human-advised action instead. The QDP framework is adapted to the twin delayed deep deterministic policy gradient algorithm (TD3) in this article for the convenience of comparison with the state-of-the-art TD3. Specifically, the human expert in the QDP-HRL considers giving advice in the case that the difference between the twin Q -networks' output exceeds the maximum difference in the current queue. Moreover, to guide the update of the critic network, the advantage loss function is developed using expert experience and agent policy, which provides the learning direction for the QDP-HRL algorithm to some extent. To verify the effectiveness of QDP-HRL, the experiments are conducted on several continuous action space tasks in the OpenAI gym environment, and the results demonstrate that QDP-HRL greatly improves learning speed and performance.

Data-driven Huntington's disease progression modelling and estimation of societal cost in the UK.

We develop a Huntington's disease (HD) progression model and integrate this with a novel economic model, accounting for the major factors of the HD's societal cost. Data from the Enroll-HD observational study were used to fit a continuous-time hidden Markov disease progression model, which identified five distinct states. The number of disease states was determined using a cross-validated maximum likelihood approach. A novel data augmentation method was used to correct the biased life expectancy of the progression model. Multiple sources of cost data were then mapped to Enroll-HD variables using expert experience. A simulation of a synthetic patient population was used to show the feasibility of the approach in estimating population costs and the impact of hypothetical intervention scenarios. Our results confirm that early cognitive decline, which is not captured by the total functional capacity score currently used by clinicians but flagged up in HD integrated staging system, can be quantified from participants' visits. Finally, the results of the UK cost modelling show that indirect costs of HD such as state benefits and lost gross domestic product contribution could be the driving factors for the societal cost, over and above health and social care costs.

Emotion-Semantic-Aware Dual Contrastive Learning for Epistemic Emotion Identification of Learner-Generated Reviews in MOOCs.

Identifying the epistemic emotions of learner-generated reviews in massive open online courses (MOOCs) can help instructors provide adaptive guidance and interventions for learners. The epistemic emotion identification task is a fine-grained identification task that contains multiple categories of emotions arising during the learning process. Previous studies only consider emotional or semantic information within the review texts alone, which leads to insufficient feature representation. In addition, some categories of epistemic emotions are ambiguously distributed in feature space, making them hard to be distinguished. In this article, we present an emotion-semantic-aware dual contrastive learning (ES-DCL) approach to tackle these issues. In order to learn sufficient feature representation, implicit semantic features and human-interpretable emotional features are, respectively, extracted from two different views to form complementary emotional-semantic features. On this basis, by leveraging the experience of domain experts and the input emotional-semantic features, two types of contrastive losses (label contrastive loss and feature contrastive loss) are formulated. They are designed to train the discriminative distribution of emotional-semantic features in the sample space and to solve the anisotropy problem between different categories of epistemic emotions. The proposed ES-DCL is compared with 11 other baseline models on four different disciplinary MOOCs review datasets. Extensive experimental results show that our approach improves the performance of epistemic emotion identification, and significantly outperforms state-of-the-art deep learning-based methods in learning more discriminative sentence representations.